Statement of Isabel V. Sawhill, Senior Fellow, Brookings Institution,
and President, National Campaign to Prevent Teen Pregnancy

Testimony Before the Subcommittee on Human Resources
of the House Committee on Ways and Means

Hearing on Welfare Reform Reauthorization Proposals

April 11, 2002

Chairman Herger, Ranking Member Cardin, and Members of the Committee:

I am pleased to have the opportunity to testify today on proposals to reauthorize the 1996 welfare reform law. I serve as a Co-Director of the Brookings Institution’s Welfare Reform and Beyond Initiative, and as part of that effort we have carefully reviewed and synthesized a very large volume of research, have talked with many state and local officials as well as other interested “stakeholders,” and have done some analysis of different proposals to encourage work or strengthen families. I also serve (part-time and on a volunteer basis) as President of the National Campaign to Prevent Teen Pregnancy, a nongovernmental organization chaired by former Governor Tom Kean. I should emphasize, however, that my testimony today reflects my own views and not the views of any organization with which I am affiliated.

Our work at Brookings has convinced me that welfare reform has been much more successful than many people anticipated. Some of this success is the result of the robust economy that prevailed in the late 1990s and to the expansion of work supports such as the Earned Income Tax Credit. But much of the success we have had in reducing caseloads, increasing employment among single mothers, and lowering child poverty must be attributed to the 1996 law. In reauthorizing the law, I believe we can build on that success. In doing so, I want to suggest that Congress give particular attention to the following: keeping the focus on moving people into unsubsidized jobs rather than placing them in government-funded work slots, making work pay, breaking the cycle of poverty by investing in child care and early childhood education, and increasing the proportion of children being born to, and raised by, two-parent, married families.  Since my time is limited, and these are large topics, I will focus the remainder of my testimony on the last objective.

Strategies for Reducing the Growth of Single Parent Families 

Most people would agree that the ultimate goal is to increase the number of children growing up with two involved parents. Three strategies for doing so are currently under discussion: reducing divorce (or improving relationships) by providing marriage counseling or education to existing couples or those contemplating marriage, helping unwed fathers to support their children and/or to marry their child’s mother,  and reducing out-of-wedlock childbearing, especially among teens. These agendas are not necessarily mutually exclusive, but they involve different strategies and different target groups (the already married or about-to-be married, the unmarried who have children, and the unmarried who don’t have children). In what follows, I want to argue that marriage is a good thing but that preventing early childbearing among those who are young and unmarried but at high risk of becoming unwed mothers and ending up on welfare is likely to be a particularly effective strategy for achieving this goal.  (Note that roughly half of all mothers on welfare had their first baby as an unmarried teenager.)   

Reducing divorce rates can contribute to fewer children being raised in single parent families. However, after increasing sharply in the 1960s and 1970s, divorce rates have leveled off or even declined modestly since the early 1980s. Moreover, children in divorced families more often retain a relationship with both parents, are more likely to receive support from a nonresident father, are less likely to need, and receive, welfare or other government assistance, and are generally much better off than those born to never-married mothers. Finally, virtually all of the increase in child poverty between 1980 and 1996 was related to the  increase in nonmarital childbearing over this period, not to greater divorce. In short,  efforts to strengthen marriages in ways that reduce the likelihood of divorce should be welcomed but divorce rates, though high, are not the crux of the problem and thus arguably should not be the focus of any new effort.

The much bigger problem is too many unmarried women having babies. Most of these women are very young when they have their first child. While only 30 percent of all nonmarital births are to women under the age of 20, half of first nonmarital births are to teenagers and most of the rest are to women in their early twenties. ([i]) So, the pattern typically begins in the teenage years or just beyond, but once begun often leads to additional births outside of marriage. There are two solutions to this problem. One is to encourage these young women to marry the fathers of their children (assuming the fathers are willing). The other is to get them to delay childbearing until they are older and married.

As Chart 1 shows, most women eventually do marry (90 percent by age 45). The problem is one of timing. Up until their mid-twenties, more women have had babies than have ever been married. But after that age, the reverse is true: the number of  women who have ever married exceeds the number who have ever had a child. So those calling for more marriage are really calling for earlier marriages.  The drawback of this solution is that it requires reversing a strong and generally healthy trend toward later age at first marriage among both men and women.  Between 1960 and 1999, age at first marriage increased from 20 to 25 for women and from 23 to 27 for men. Age at first marriage is one of the strongest predictors of marital stability and this trend toward later marriage is a very important – probably the single most important – reason for recent declines in the incidence of divorce. One recent study by Tim Heaton  at Brigham Young University based on data from the National Survey of Family Growth finds that all of the decline in divorce rates since 1975 is related to the increase in age at first marriage. ([ii]) Not only is this trend good for marriage, it is good for children as well. Younger mothers often lack the maturity, patience, and education that have been shown to produce better outcomes for children.

The argument will be made that in earlier times it was common for women to marry young. But our economy now demands much more education than in earlier periods and provides women as well as men an opportunity to pursue both education and a career beyond high school. To be sure, some women may want to forego such opportunities in order to become full-time wives and mothers at an early age; but a social policy that actively encourages such early marriage would be inconsistent with one that also sees investments in education and in stable long-term marriages as socially beneficial.

Perhaps what is really intended by marriage advocates is not a set of policies that would encourage earlier marriages across the board but only in cases where a woman is already pregnant or has had a child. Such “shotgun” or “after-the-fact” marriages to the biological father were common in the past but have virtually disappeared in recent years. Their modern counterpart is what is often called fragile family initiatives – efforts to work with young couples, many of whom are romantically involved or cohabiting at the time of the baby’s birth, to help them form more stable ties and where appropriate, marry. These efforts often involve education, training, counseling, and peer support for the fathers. An evaluation of one such effort, Parents Fair Share, produced somewhat disappointing results. ([iii]) But it would be premature to write off such efforts. About two-fifths of all out-of-wedlock births are to cohabiting couples and cohabitation seems to be rapidly replacing marriage as a preferred living arrangement among the younger generation. These cohabiting families are much less stable than married families. Less than half of them stay together for five years or more. ([iv]) Whether such couples can be persuaded to marry and whether these marriages would endure if they did is not entirely clear, but some research suggests that marriages preceded by cohabitation are less stable than those that are not. ([v]) In the meantime, any program that provides special supports, such as education and training, to unwed parents, whether mothers or fathers, runs the risk of rewarding a behavior that society presumably would like to discourage. 

Many unwed mothers cohabit not with the biological father of their children but with another man and some of these relationships may also end in marriage. But, surprising as it may seem, such stepfamilies seem to be no better for children than being raised in a single parent home.

More importantly, once a woman has had a child outside of marriage, her chances of marrying plummet. Daniel Lichter of the Ohio State University finds that the likelihood that a woman of a given age, race, and socioeconomic status will be married is almost 40% lower for those who first had a child out of wedlock (and 51% lower if we exclude women who marry the biological father within the first 6 months after the birth). By age 35, only 70 percent of all unwed mothers are married in contrast to 88 percent among those who have not had a child. He compares women who had a premarital pregnancy terminated by a miscarriage to those who carried to term, and finds that these differences in marriage rates persist. ([vi]) This suggests that having a baby out of wedlock causes women to marry less rather than simply reflecting the pre-existing characteristics of this group of women. The reasons unwed mothers are less likely to marry are unclear. They may be less desirable marriage partners, may be less likely to spend time at work or in school where they can meet marriageable men, or may simply lose interest in  marriage once they have children. Moreover, having had one child out of wedlock, they appear to be relatively uninhibited about having additional children in the same way. In short, early unwed childbearing leads to less marriage and more illegitimacy. Thus, one clear strategy for bringing back marriage is to prevent the initial birth that makes a single woman less marriageable throughout her adult years. Most young women aspire to marry and publicizing their much reduced chances of marrying once they have a baby might make them think twice about becoming unwed mothers.

Not only are unwed mothers less likely to marry than those without children but when they do marry, they do not marry as well. Their partners are more likely to be high school dropouts or unemployed than the partners of women who have similarly disadvantaged backgrounds but no children. Although marriage improves on unwed mothers’ chances of escaping from poverty, it does not offset the negative effects associated with an unwed birth, according to Daniel Lichter and his colleagues. ([vii])

My conclusion is that efforts to promote marriage and reduce divorce hold little promise for curbing the growth of single parent families and that what is needed instead is a serious effort to reduce early, out-of-wedlock childbearing. Moreover, as I will argue shortly, unlike encouraging marriage, this is something we actually know how to do. And finally, although some of what needs to be done is controversial, it is no more so than the promarriage agenda that some now tout. According to the Pew Research Center for the People and the Press, the American public is not in favor of the government developing programs that encourage people to get and stay married. Indeed, 79% prefer that the government “stay out” of such activities. Only 18% favor the idea. The group most in favor of this agenda is highly committed white evangelicals but only 35% of this subgroup favors government involvement in encouraging marriage while 60% remain opposed.

Let me be clear that I am not arguing against marriage as a social goal. I am arguing that the most effective and least controversial way to accomplish this goal is to insure that more young women reach the normal age of marriage having finished school, established themselves in the workplace, and done both without having borne a child. The chances that they will then have children within marriage, that the marriage will be a lasting one, and that their children will receive good parenting will be much greater. The chances of achieving this goal will be enhanced if the message young people receive from society is not just that delaying parenthood is important, but also that children belong within marriage. As Wade Horn notes, too many teen pregnancy prevention programs have left the impression that it’s fine to have a baby without being married as long as you wait until you’re age 20. ([viii]) But of course there is nothing magic about leaving the teen years. What needs to be stressed instead is accomplishing various life tasks, such as completing one’s education and finding a lifetime partner before becoming a parent. Young people accomplish these tasks at different ages but few are ready before their early twenties at best.  

None of this is meant to imply that it is not worthwhile to use the bully pulpit to restore a marriage culture, provide pre-marital education and counseling, and engage faith-based communities, schools, and parents in sending different messages to young people about the benefits of marriage. In addition, attention should be given to some of the financial disincentives to marriage, especially in low-income communities. Congress acted in 2001 to reduce the marriage penalty in the tax code, including the large marriage penalty associated with the EITC. And many states have liberalized welfare eligibility standards for two parent families. More could be done but any meaningful reduction of marriage penalties in income-tested programs carries enormous budgetary costs and is unlikely to have more than small effects on behavior. So, without a strong effort to prevent early childbearing, I very much doubt that these efforts alone will significantly reduce the growth of single parent families and improve economic and social environments for children.

Reducing Early Childbearing

After climbing steadily at almost 1 percentage point per year for over twenty years, the proportion of all children born outside of marriage (“the nonmarital birth ratio”) leveled off after 1994. Much of the good news is related to a decline since 1991 in the teenage birth rate. (Almost four out of every five teen births is out-of-wedlock.). In fact, as Chart 2 shows, if there had been no decline in the teen birth rate, the nonmarital birth ratio would have continued to climb in the late 1990s, albeit not as rapidly as in the prior decade. More specifically, if teen birth rates had held at the levels reached in the early 1990s, by 1999 the nonmarital birth ratio would have been more than a percentage point higher. This suggests that a focus on teenagers (although not to the exclusion of women in their early twenties who also contribute disproportionately to these trends) has a major role to play in reducing both out-of-wedlock childbearing and the growth of single parent families.

This conclusion is reinforced when one recalls that teens who avoid a first nonmarital birth are more likely to marry and less likely to have additional children outside of marriage. These indirect effects are not included in Chart 2, but as noted above, they are likely to be substantial. 

Since the decline in the teenage birth rate has contributed significantly to the leveling off of the nonmarital birth ratio, it is worth asking what caused the decline and whether further steps can be taken to lower the rate (and ratio) further.

Teen births are down because teen pregnancies are down. (The difference between them depends on how many teens have an abortion, and after increasing in the decade immediately following Roe v. Wade, abortion rates for teens, as for all women, have now leveled off or declined.) The decline in teen pregnancy rates has been driven, in turn, by both declining rates of sexual activity among teens and better contraception. Proponents of abstinence like to think that the former has been most important while proponents of birth control give greater weight to changes in contraceptive behavior. With existing data, it’s not possible to determine the precise role of each, but almost everyone agrees that both have played a role. ([ix]) That said, there is a growing public consensus that abstinence is preferable, especially for school-age youth, but that contraception should be available. Polling by the National Campaign to Prevent Teen Pregnancy has consistently found majority support for this view with 73 percent of adults agreeing with the proposition that teens should not be sexually active but that teens who are should have access to contraception. Support for this moderate position has increased 14 percent since 1996. ([x])

These data on reduced sexual activity suggest that the emphasis on abstinence, including new funding for abstinence education in the 1996 welfare reform bill, is working to reduce teen pregnancies and out-of-wedlock births. Yet evaluations of abstinence education programs have thus far failed to show much evidence of success. My conclusion is that new messages about abstinence are having an impact but less because they are embedded in so-called “abstinence only” education programs and more because they have infected the entire culture including traditional sex education programs, the media, faith-based efforts, and the way in which parents communicate with their children. The abstinence message is no longer the exclusive province of a small band of conservative activists; it is now being promoted by many organized groups (including the National Campaign to Prevent Teen Pregnancy) and is widely endorsed by most ordinary Americans including parents, teachers, many political leaders, and to a lesser degree, by teens themselves. This shift in both attitudes and behavior during the 1990s is significant and has clearly contributed to the decline in teen and out of wedlock childbearing. ([xi])

Other factors that may have played a role include fear of AIDS and other sexually transmitted diseases in combination with more, or more effective, sex education programs (discussed in more detail below). Finally, welfare reform itself in combination with a strong economy may have had an impact. Although the decline in teen pregnancy and birth rates predates welfare reform, most of the decline prior to 1996 was the result of a drop in second or higher order births to teens who were already mothers and appears to have been caused by the availability for the first time of longer-lasting,  more effective forms of contraception such as Depo Provera. These methods are not widely used but have caught on particularly among the subgroup of young women who have already had a baby. It was not until the latter half of the 1990s that first births to teens began to decline significantly. ([xii]) Whether this decline in first births is the result of welfare reform or not is uncertain; but it needs to be emphasized that the 1996 law sent a new message not only to young women but also to young men. The message to young women was financial support for you and your baby is going to be time limited and require that you work. The message to young men was if you father a child, you will be responsible for its support. And several studies have found that tougher child support enforcement reduces out-of-wedlock childbearing. ([xiii]) Thus, the evidence is at least consistent with the view that welfare reform has played a role in producing the observed trends.

Building on Success

Other data reinforces the view that welfare reform may be affecting family formation. Not only has the teen birth rate declined and the nonmarital birth ratio leveled off, but in the late 1990s the proportion of children living in a single parent family stabilized or even declined modestly for the first time in many decades. ([xiv]) This reversal of trend was most notable for low-income families, and those with less education or very young children, just as one would expect if welfare reform were the cause. Looking at data for 1997 and 1999, for example, Gregory Acs and Sandi Nelson of the Urban Institute find that the share of families composed of single mothers living independently declined almost 3 percentage points more among families in the bottom income quartile than among those in the second quartile. ([xv])

Changes in such behaviors as divorce and out-of-wedlock childbearing are likely to respond only slowly to a shift in the policy environment and it would be premature to attribute all or even most of these changes to the 1996 law. But it would also be wrong, in my view, to say that it has not had an effect simply because evaluations of some of the specific provisions such as family caps or the illegitimacy bonus or abstinence education programs have not shown clear impacts. ([xvi]) Arguably, much more important than any of these are new messages about time limits, about work, and about abstinence. Young women who decide to have children outside of marriage now know that they will receive much more limited assistance from the government and that they will be expected to become self-supporting. Young men are getting the message that if you father a child you will be expected to pay child support. Teenagers who choose to remain abstinent now feel much more support from program operators, advocates, and peers. If I am right about this, then one important recommendation for policy makers is that they maintain the current thrust of the law. However, programmatic micromanagement of various family behaviors at the federal level is another matter. Detailed prescriptions about how funds can be used at the local level are likely to be neither effective nor widely supported. Broader messages about work, about family formation, about abstinence, and about the need for fathers to support their children should be sufficient.

The main actors in this story are not the federal government but states, communities, and nonprofit (including faith-based) organizations. And what they need are resources, technical assistance, and information about what might work to reduce early childbearing outside of marriage and slow the growth of single parent families. Current efforts are fragmented, underfunded, and often ineffective. For all of the reasons stated earlier, the focus needs to be on reaching young people before they have children. The high-risk group includes not only teenagers but also those in their early twenties. But attitudes about sex, relationships, and marriage are formed at an early age and the intense interest in them that develops during the adolescent years produces an especially receptive audience at this time.

The good news is that in the past five years, research on teen pregnancy prevention programs has found a number that work. Douglas Kirby’s review, Emerging Answers, published in the summer of 2001, identifies several rigorously evaluated programs that have reduced teen pregnancy rates by as much as one half. ([xvii]) Some effective programs involve teens in community service or afterschool activities with adult supervision and counseling. Others focus more on  sex education but not necessarily just on teaching reproductive biology. The most effective sex education programs provide clear messages about the importance of abstaining from sex or using contraception, teach teens how to deal with peer pressure, and provide practice in communicating and negotiating with partners.  This research needs to be aggressively disseminated so that local efforts are based on more informed judgments. And since there are a variety of different approaches that can be effective, communities should be allowed to choose from among them based on their own needs and values.   Simultaneously, much more emphasis needs to be placed on the potential of sophisticated media campaigns to change the wider culture. Such campaigns have been used to effectively change a variety of health behaviors in the past but their full potential has not been tapped in this arena. ([xviii]) Some nonprofit groups, such as the National Campaign to Prevent Teen Pregnancy and the National Fatherhood Initiative, are working in partnership with the media to embed new messages into the television shows most often watched by teens. And many states are using the abstinence education funds from the welfare reform bill for public service announcements, but additional resources, including some that could be used to design and implement a national effort, are needed.

Conclusion     

The goal of increasing marriage, is, in my view, entirely laudable. However, it needs to be reconciled with other goals, such as supporting children who are already born. One extreme option would be to eliminate benefits entirely for those living in single parent families or for young women who bear a child out of wedlock.  A softer version of this would be to earmark some portion of existing government benefits for those who are married or to carve out a portion of the welfare dollars that go to the states for marriage education or other pro-marriage activities.

These policies would come on top of the reforms instituted in 1996 which sent a strong message that women who bear a child outside marriage will no longer be able to raise that child without working and that the men who father such children will have to contribute to their support. The early indications are that these messages may be having an effect: teen birth rates have fallen, the share of children born out of wedlock has leveled off, and the share of young children living in married families have all increased in the late 1990s.

These developments suggest that current policies may be working, and given time for new social norms to evolve, will have larger effects. Pushing pro-marriage policies to the next level could upset the fragile political coalition supporting current reforms. Liberal advocates argue that such proposals effectively divert resources away from helping single parents raise their children. Whatever mistakes the parents may have made, few people want to deprive their children of assistance as a consequence.

The key behavior here is not marriage per se but childbearing outside of marriage. Divorce rates may be high but they are not increasing and have played no role in the growth of single parent families for several decades. Virtually all of that growth, and the associated growth in child poverty in the 1980s and early 1990s, was caused by increased childbearing among young, single women. Moreover, half of that childbearing begins in the teenage years and most of the rest of it takes place among women in their early twenties. Once such women have had a child their odds of ever getting married plummet. In fact, having established a single parent household, these women often go on to have a second or third child, often with different fathers. Many point to the shortage of “marriageable men” – that is, men with good job prospects -- in the communities where these women live; but there is a shortage of “marriageable women” as well. Most men are going to think twice about taking on the burden of supporting someone else’s child.

There are only two solutions to the problem of childbearing outside of marriage. One is to encourage young women to marry very young, say in their teens or their early twenties at the latest, before they start having children. The other is to persuade them to delay childbearing until they are in their mid-twenties.  Although commonplace as recently as the 1950s, early marriage is no longer a sensible strategy in an economy where decent jobs increasingly require a high level of education and young people need to spend the first few years out of school getting established in the job market. Moreover,  teen marriages are twice as likely to end in divorce as marriages among adult women in their mid-twenties. ([xix]) So if we want to encourage marriage, prevent divorce, and ensure that more children grow up with married parents, we must first insure that more women reach adulthood before they have children. It is a necessary if not sufficient condition for success. It implies redoubling efforts to prevent teen pregnancy. These efforts have now been carefully evaluated and many of them appear to be quite effective.

So-called fatherhood programs which work directly with young men may also help but so far such efforts do not have a solid track record of success and send the wrong message if resources are targeted only on men who have already fathered a child out of wedlock. A far more promising strategy is to focus on young men and women who have not yet had a baby, to convince them there is much to lose if they enter parenthood prematurely, and much to gain if they wait until they are married.

  [i] National Center for Health Statistics, “Births: Final Data for 1999,” National Vital Statistics Report 49-1 (Hyattsville, MD: National Center for Health Statistics, 2001) 44.

  [ii] Tim Heaton, “Factors Contributing to Increasing Marital Stability in the United States” Brigham Young University, July 2000, 12-13.

  [iii] Virginia Knox and Cindy Redcross, Parenting and Providing: The Impact of Parents’ Fair Share on Paternal Involvement (New York: Manpower Demonstration Research Corporation, 2000).

  [iv] Elizabeth Terry-Humen, Jennifer Manlove and Kristin A. Moore, “Births Ouside of Marriage: Perceptions vs. Reality,” Child Trends Research Brief (April 2001) 4.

  [v] The National Marriage Project, “Social Indicators of Marital Health and Wellbeing,” The State of Our Unions 2001 (Piscataway, NJ: Rutgers University, 2001) 24.

  [vi] Daniel T. Lichter and Deborah Roempke Graefe, “Finding a Mate? The Marital and Cohabitation Histories of Unwed Mothers,” Out of Wedlock: Trends, Causes and Consequences of Nonmarital Fertility, eds. Lawrence L. Wu and Barbara Wolfe (New York: Russell Sage Foundation, 2001) 329.

  [vii] Daniel T. Lichter, Deborah Roempke Graefe and J. Brian Brown, “Is Marriage a Panacea? Union Formation Among Economically-Disadvantaged Unwed Mothers,” The Ohio State University, April 2001, 18-19.

  [viii] Wade F. Horn, “Confronting the ‘M’ Word,” American Experiment Quarterly 4 (2001): 85

  [ix] Christine Flanigan, What’s Behind the Good News (Washington: The National Campaign to Prevent Teen Pregnancy, 2001) 7.

  [x] The National Campaign to Prevent Teen Pregnancy, With One Voice (Washington: The National Campaign to Prevent Teen Pregnancy, 2001) 5.

  [xi] Leighton Ku, Freya L. Sonenstein, Laura D. Lindberg, Carolyn H. Bradner, Scott Boggess, and Joseph H. Pleck, "Understanding Changes in Sexual Activity among Young Metropolitan Men: 1979-1995," Family Planning Perspectives Vol. 30 (November-December 1998).

  [xii] The National Campaign to Prevent Teen Pregnancy, Just the Facts (Washington: The National Campaign to Prevent Teen Pregnancy, 2000) 100-102.

  [xiii] Irwin Garfinkel, Theresa Heintze, and Chien-Chung Huang, “Child Support Enforcement: Incentives and Well-Being,” The Incentives of Government Programs and the Well-Being of Families, eds. Bruce Meyer and Greg Duncan (Chicago: Joint Center for Poverty Research, 2000) 10.

[xiv] Richard Bavier, “Recent Increases in the Share of Young Children Living with Married Mothers,” Washington: Office of   Management and Budget, September 2001, 3.

   [xv] Gregory Acs and Sandi Nelson, “ ‘Honey, I’m Home.’ Changes in Living Arrangements in the Late 1990s,” New Federalism Urban Institute Policy Brief B-38 (June 2001) 5.

  [xvi] Charles Murray, “Family Formation,” New World of Welfare eds. Rebecca Blank and Ron Haskins (Washington: Brookings Press, 2001) 145.

  [xvii] Douglas Kirby, Emerging Answers (Washington: The National Campaign to Prevent Teen Pregnancy, 2001).

  [xviii] Leslie B. Snyder, “How Effective Are Mediated Health Campaigns?” Public Communication Campaign, eds. Ronald E. Rice and Charles K. Atkin (Thousand Oaks, CA: Sage Publications, 2000).

  [xix] Matthew D. Bramlett and William D. Mosher, “First Marriage Dissolution, Divorce, and Remarriage,” Center for Disease Control Advance Data 323 (May 2001) 5.